The five questions to ask before hiring anyone for this.

Every business in Rhode Island is getting emails right now about automation, agents, and the new tools. Most of them come from people who are selling the same thing to everyone, and a lot of those people learned it recently. Some are very good. Some will cost you a year and a lot of money before you find out which.

The pitch will not tell you which one you are talking to. The pitch is always good. The questions will, because a person who has done this work for real businesses answers them one way, and a person who has not answers them another. The difference is easy to hear once you know what to listen for.

You are not buying technology. You are buying someone's judgment about your business. Test the judgment.

Ask all five in the first meeting, before anyone shows you a screen. Write down the answers. Then ask them of us, because we would rather earn the work than assume it.

1. Have you ever run a business, or only built software for them?

The technology is the easy part. What is hard is knowing how a business actually runs: what a slow month feels like, why the office manager guards the schedule, what payroll on a Friday does to a decision. Someone who has carried that will build differently from someone who has read about it.

A good answer is specific. What they ran, for how long, what went wrong, what they would do differently. Or an honest "no, but here is who on our side has," which is fine as long as that person is in the room when decisions get made.

A bad answer pivots to technical credentials. Years of experience with tools is not years of experience with payroll. If the answer is a list of software, you have your answer.

2. What would you not recommend for us?

Anyone can list what they would build. The question is whether they can see what should not be built, and whether they will say so when it costs them a sale. This is the question that separates advisors from salespeople faster than any other.

A good answer is concrete and slightly uncomfortable. "I would not touch your scheduling yet, the problem is upstream." "Do not buy the private version, your volume does not justify it." "Half of what you asked for, your current software already does."

A bad answer is "everything is possible." It may even be true. But a person who cannot name one thing you should not do is either not thinking about your business or not willing to lose the sale over it.

3. Where does our data go, and who can see it?

Your customer list, your pricing, your employee records and your email are going to pass through whatever gets built. You are entitled to a plain answer about where they sit, who can look, and what the vendor's own tools do with them. Not a reassuring sentence. A location and a name.

A good answer is plain, specific, and unbothered by the question. Which systems hold what. Whether it is on your hardware, their platform, or inside tools you already use. Who on their team can see it and why. What happens to it if you part ways. If any of it is regulated, they bring that up before you do.

A bad answer is "it is all secure." Or "industry standard." Or "don't worry about that." Any answer that is a feeling rather than a fact. Watch also for someone who has clearly never been asked, because that means none of their other clients asked either.

4. What happens if we stop working together?

Every relationship ends eventually, on good terms or bad. The question is whether your business keeps running and whether your information comes with you. A surprising number of arrangements are built so that the answer is no, and nobody mentions it until the day it matters.

A good answer: your data is yours and you take it with you. What you paid to build is documented and either yours outright or available to buy out. They can tell you, today, what the hand-off would look like, because they have thought about it.

A bad answer is silence, or a laugh, or "why would you want to?" A vendor who has not planned for the end of the relationship has built something you cannot leave. That is not a partnership. It is a lease with no exit.

5. Can you show me something you built that is running right now?

Not a demo. Not a slide. Something a real company is using today, with a real result, described plainly. This is the one question a person who has only read about this cannot answer, and it is fair to ask it early.

A good answer is a story with a before and an after. "A shop was quoting one job a week. Now it is about twenty. Same estimator." They describe the problem, what they built, and what changed, in numbers. They may not name the client, and that is right. They can describe the business type and the outcome.

A bad answer is only mockups. Only "we are working with a client on that now." Only results that belong to the software vendor rather than to them. Or numbers that sound too round to be real, because they are.

One more sign, worth more than all five

Did they ask you more questions than you asked them? The people worth hiring spend the first meeting trying to understand your business. The others spend it trying to be understood.

Use these on us too

We wrote the list, so it is only fair. Ask us all five in the first thirty minutes and we will answer every one plainly, including the ones about what we would not build for you. If we are not the right fit, we will tell you who might be.

The printable version, with four more questions and a scorecard to bring to the meeting, is our guide What to Ask Before Hiring Anyone for This. Free, about ten minutes.